Pelvic osteomyelitis in children is an uncommon lesion. Pelvic osteomyelitis should be considered in the differential diagnosis in any febrile child presenting with limitation of motion around the thigh or hip and/or with gait disturbance (1). The differential diagnosis of pelvic osteomyelitis in children includes septic arthritis of the hip or sacroiliac joint, Legg-Calve-Perthes disease, toxic synovitis of the hip, and less commonly, collagen-vascular diseases, neoplasm’s involving bone and retroperitonial abscess. The diagnosis of pelvic osteomyelytis is made on the basis of clinical, laboratory and imaging examinations (5). Plain x-ray changes in acute haematogenous osteomyelitis lag between 7 to 10 days behind the evolution of infection, which are well known. The three-phase bone scan is the gold standard and it includes assessment of blood flow (angiographic phase), early uptake by bone (blood pool phase), and late uptake (delayed bone phase, 2 hours later). This permits differentiation of cellulitis from osteomyelitis. The scintigraphic appearance of osteomyelitis is increased uptake in all three phases, especially the third. A photopenic spot (cold) may be encountered in infants and is related to severe medullary thrombosis and compression of the nutrient arteries by high pressure intraosseous pus. Reported sensitivities of bone scans for detection of osteomyelitis vary from 32% to 100% (5). Using the three-phase technique and a high-resolution gamma camera, it has been found by Ferley et al. to have a sensitivity of 89% and specificity of 94% in a large series of children. Gallium–67 citrate was the first scintigraphic agent used to detect inflammatory processes. Despite initially promising results, sequential technetium–gallium scanning has demonstrated an accuracy rate of only 70%. Sequential technetium-gallium scanning is especially unreliable in patients who have another lesion, such as non-union, surgery or a neuropathic joint, at the site of suspected osteomyelitis (5). Indium-111-labelled leukocytes, indium-labelled polyclonal immunoglobulins and technetium-99-m-labelled leukocytes are now being evaluated. In a study of 100 patients with a variety of suspected infections and inflammation, a sensitivity of 100% and a specificity of 95% were reported using Tc-99m-labelled leukocytes (HMPAO-WBC). Marrow abnormality on MRI is a more sensitive indicator of disease than lytic changes seen radiographically and appears much earlier in the course of the disease. Active osteomyelitis shows a low signal on T1-weighted images and appears as a high signal on T2-weighted images (5). This pattern represents a replacement of the marrow fat with water secondary to oedema, exudates, hyperaemia and ischaemia. In a study by Morrison et al, demonstrated an improved sensitivity (88%) and specificity (93%) with fat suppressed contract enhanced MR imaging compared to unenhanced MR imaging (sensitivity of 79% and specificity of 53%). The intravenous contrast agent, gadopentetate di-N-methylglucamine, a paramagnetic material, may also be useful in differentiating vascularized and inflamed tissue from the peripheral rim enhancement characteristic of an abscess. The MRI technique appears promising in the diagnostic work-up of osteomyelitis, but remains of limited use for whole body examination. This is important particularly in the paediatric age group where multifocal disease is seen in 7% of cases. Scintigraphy is preferred to MRI for the initial evaluation, because it is less expensive, rarely requires sedation and children can be scanned more than once, thus improving the detection rate. MRI is reserved for cases that require surgical intervention such as infections in the spine and pelvis, and difficult cases in which there is conflicting clinical data or infection that fails to respond to antibiotic treatment. A comparative study by Unger et al quoted sensitivities of MR and scintigraphy for diagnosis of osteomyelitis were 92% and 82% and the specificities were of 96% and 65% respectively. Because of its ability to separate soft tissue disease from underlying bone marrow, MR may be used to evaluate patients with positive bone scintigraphy to improve the specificity and accuracy of diagnosis for osteomyelitis. CT should be used as a second or confirmatory study when a single nuclear scintigram is normal or equivocal and pelvic osteomyelitis is strongly suspected on clinical grounds. Spiral CT especially is superior to MRI for visualising bony destruction, gas in the bone and bony sequestration and is useful in the spine, pelvis and sternum. Ultrasound scanning is a non-invasive, easily available technique, which does not use ionising radiation; in subperiosteal abscesses, ultrasound-guided aspiration and drainage may prove useful in some cases. In a review of currently used imaging techniques for osteomyelitis Sammak et al. concluded that various techniques are important and complimentary, rather than competitive, in the diagnosis of osteomyelitis. It is vital to understand the limitations of each imaging modality in order to avoid any delay in the diagnosis and management and prevent possible complications.
